Optimization of acquisition and contrast injection protocol for C-arm CT imaging in transcatheter aortic valve implantation: initial experience in a swine model.

To determine the optimal C-arm computed tomography (CT) protocol for transcatheter aortic valve implantation (TAVI) in swine. In 6 swine, C-arm CT was performed using 5-s ungated acquisition during sinus rhythm with aortic root (Method 1) or peripheral (Method 2) injection, and during rapid ventricular pacing with root injection (Method 3). Additionally, 24-s ECG-gated acquisitions were performed during sinus rhythm with root (Method 4) or peripheral (Method 5) injection. Aortic root enhancement, presence of artifacts and contrast volumes were compared for all methods. Aortic root measurements were also compared between C-arm CT and multidetector-row computed tomography (MDCT). The best C-arm CT image set was identified and used to predict optimal angiographic projection angles during TAVI; predictions were compared to those from MDCT. Methods 1, 3, 4, and 5 yielded sufficient root enhancement with mild or moderate artifacts and aortic annulus, sinotubular junction, and mid-ascending aorta diameters similar to MDCT. Ungated C-arm CT (Methods 1, 3) required less contrast than ECG-gated C-arm CT (Methods 4, 5). Method 3 was optimal yielding images with high attenuation, few artifacts (2.0), and root measurements similar to MDCT using minimal contrast (36 mL). Predicted angiographic projections from Method 3 were similar to MDCT. Ungated C-arm CT during rapid pacing with aortic root injection required minimal contrast, yielded high attenuation and few artifacts, and aortic root measurements and predicted angiographic planes similar to those from MDCT.